Discussion forums provide an increasingly popular medium through which people may participate in an information sharing community. In a discussion forum people may ask questions, share opinions, provide answers to questions, and so on. Discussion forums typically employ an electronic bulletin board to provide access to various “conversations” or “threads”. As the number of conversations increases and as the number of messages added to the increasing number of conversations expands it can be difficult for users to locate and/or isolate content in which they are interested. Thus, users may be overwhelmed by the volume of information they may need to navigate to locate specific information. Additionally, not all the information may be relevant to the user and not all the information may have equal value. Thus, a user may become distracted by irrelevant and/or useless threads, lose focus, and become lost.
Relevancy in the context of a discussion forum may be regarded from different points of view. A first aspect of relevancy may be traditional keyword based relevancy. A user may be looking for information in threads (e.g., conversations) that discuss a certain subject. A conversation may include an original message and replies to that message. Conventionally, these threads and/or messages within the threads may be found using a keyword search. However, this type of search may provide a large number of threads and/or messages that are potentially irrelevant. Some discussion forums therefore provide additional information about a thread and/or a message. For example, a discussion forum may consider information about how many times a certain thread has been accessed and/or how many times a particular message in a thread has been accessed. This information may be used to further rank items retrieved in response to a keyword based search. Additionally and/or alternatively, this information may be provided to users so that they can decide which threads and/or messages to review. Some discussion forums may also provide other statistics (e.g., reply count, user rating) that may be presented to a user and/or used to further rank items previously ranked by keyword relevancy.